Project:Tukituki Streamflow Depletion Assessment

(1)

Project:Tukituki Streamflow Depletion Assessment

Draft B

7 October 2013

(2)

The SKM logo trade mark is a registered trade mark of Sinclair Knight Merz Pty Ltd.

Project: Tukituki Streamflow Depletion Assessment

Draft B

7 October 2013

Sinclair Knight Merz Carlaw Park

12-16 Nicholls Lane, Parnell Auckland 1010, New Zealand T +64 9 928 5500

F +64 9 928 5501 www.globalskm.com

COPYRIGHT: The concepts and information contained in this document are the property of Sinclair Knight Merz Limited (SKM). Use or copying of this document in whole or in part without the written permission of SKM constitutes an infringement of copyright.

LIMITATION: This report has been prepared on behalf of and for the exclusive use of SKM’s client, and is subject to and issued in connection with the provisions of the agreement between SKM and its client. SKM accepts no liability or responsibility whatsoever for or in respect of any use of or reliance upon this report by any third party.

(3)

List of Abbreviations 1

Glossary of Terms 2

Executive Summary 4

1. Introduction 6

1.1. Purpose and Scope of Report 6

2. Tukituki Physical Setting 7

2.1. Surface Water Catchments 7

2.2. Geology 7

2.3. Hydrogeology 9

2.4. Groundwater Abstraction 11

3. Plan Change 6 12

3.1. Publically Notified Version of TT11 12

3.2. Policy TT11 Version Submitted by Ms Codlin (HRBC) 13

3.3. Proposed Amendments to Policy TT11 14

3.4. Version adopted for this assessment 15

4. Methodology 16

4.1. Overview 16

4.2. SKM Stream Depletion Calculations 16

4.3. HBRC Stream Depletion Calculations 17

4.4. Data Sources 18

4.5. Input parameters 18

4.5.1. Transmissivity and Aquifer Storage 18

4.5.2. Distance to the River 19

4.5.3. Seasonal Allocation 19

4.5.4. Aquifer Confinement 20

4.5.5. Multiple Wells per Consent 20

4.6. SKM Modelled Scenarios 20

4.7. Comparison to Plan Change Table 5.9.7 Criteria 20

4.8. HBRC Streamflow Depletion Modelling 21

5. Results 22

5.1. Scenario 1: SKM - Calculating artesian and unconfined separately and

L = <1 L/s 22

5.2. Scenario 2: SKM - Calculating artesian and unconfined separately and 23

(4)

PAGE ii

5.4. Scenario 4: HBRC - Hunt (2003), Low = <1 L/s 24 5.5. Scenario 5: HBRC - Jenkins, Low = <2 L/s 24 5.6. Scenario 6: HBRC – Hunt (2003), Low = <2 L/s 25

6. Discussion 26

6.1. Aquifer Conceptualisation 26

6.2. Calculation of Seasonal Allocation 27

6.3. Aquifer Parameters 28

7. References 29

(5)

Document history and status

Revision Date issued Reviewed by Approved by Date approved Revision type Draft A 3 October 2013 Gillian Holmes Gillian Holmes 3 October 2013 Internal QA Draft B 4 October 2013 Gillian Holmes Michelle Sands 4 October 2013 Final QA Review

Distribution of copies

Revision Copy no Quantity Issued to

Draft B 1 1 Horticulture NZ

Printed: 7 October 2013

Last saved: 7 October 2013 05:10 PM

File name: Tukituki Catchment Streamflow Depletion Assessment Author: Tim Baker and Catherine Sturgeon

Project manager: Nic Conland Name of organisation: Horticulture NZ Name of project: Tukituki Plan Change 6

Name of document: Tukituki Catchment Streamflow Depletion Assessment Document version: Draft B

Project number: AE04485

(6)

I:\AENVW\Projects\AE04485\Deliverables\Reports\Groundwater\AE04485-NCM-RP-0001_Groundwater Connectivity_RevB.docx PAGE 1

List of Abbreviations

Abbreviation Full

(7)

Glossary of Terms

Term Definition

Aquiclude A geologic formation, group of formations, or part of a formation through which virtually no water moves.

Aquifer A formation, group of formations, or part of a formations that contains sufficient saturated permeable material to yield economical quantities of water to wells and springs.

Aquitard

A saturated, but poorly permable bed, formation, or group of formations that does not yield water freely to a well or spring.

However, an aquitard may transmit appreciable water to or from adjacent aquifers.

Artesian A well deriving its water from a confined aquifer in which the water level stands above the ground surface; synonymous with flowing artesian.

Confined

A formation in which the groundwater is isolated from the atmosphere at the point of dischagre by impermeable geologic formations; confined groundwater is generally subject to pressure greater than atmosphere.

Connectivity The degree of the hydraulic connection between a stream and the groundwater source for a well.

Hydraulic gradient The rate of change in total head per unit of distance of flow in a given direction.

Leakage The transmission of water between aquifers separated by a low permeability layer or aquitard.

Piezometric contours An imaginery surface representing the total head of groundwater in a confined aquifer that is defined by the level to which water will rise in a well.

Semi-confined An aquifer that is partly confined by layers of

low permeability material through which recharge and discharge may occur.

Specific yield The ratio of the volume of water that a given body of rock or soil will

(8)

Storage co-efficient The volume of water an aquifer releases from or takes into storage per unit surface area of the aquifer per unit change in head.

Stream depletion factor

A term used to quantify the relation between the distance of a pumping well from a nearby stream and the hydraulic diffusivity of an aquifer.

Streamflow depletion A decrease in stream flow due to the effect of groundwater pumping.

Transmissivity

The rate at which water is transmitted through a unit width of an aquifer under a unit hydraulic gradient. It is given in cubic meters per day through a vertical section of an aquifer one meter wide and extending the full saturated height of an aquifer under a hydraulic gradient of 1.

(9)

Executive Summary

Tukituki River Catchment Plan Change 6 (PC6) is a catchment-specific change to the Hawke's Bay Regional Resource Management Plan. Under Policy TT11 of PC6 groundwater consent holders will be required to assess their degree of connectivity to the Tukituki River (or tributaries). Depending on the outcome of their assessment (compared against Table 5.9.7 – Low, Medium, High or Direct) the portion of the take that comes from the river can be counted as surface water allocation.

There are a total of 134 boreholes in the Lower Tukituki, and 113 groundwater take consents. The majority of these groundwater take consents are for cropping, orchards and vineyards. These are all land uses that are reliant on irrigation water for business continuity, and cumulatively provide a significant economic contribution to the region.

SKM and HBRC have both completed high level modelling exercises to estimate the potential streamflow depletion resulting from the currently consented groundwater abstraction in the Lower Tukituki Surface Water Allocation Zone.

A total of 312 L/s of streamflow depletion is allowed for under Schedule XVIII of Change 6. This 312 L/s comes from totalling the instantaneous abstraction rate of all groundwater takes currently assessed by HBRC as Stream Depleting. No reassessment of these takes was undertaken as part of the Plan Change 6 development process, and the methodology initially used for assessing these takes is often inconsistent with the methodology that is proposed in Policy TT11 of PC6.

A high level modelling exercise was completed by SKM using the Jenkins Streamflow depletion methodology. The results of this modelling showed a potential for 588 L/s of streamflow depletion in the Lower Tukituki. This figure assumes that the threshold for the ‘Low’ category of Table 5.9.7 is changed to <2 L/s, as proposed in the evidence of Mr Michael Thorley.

Alternative modelling undertaken by HBRC using the Hunt (2003) methodology and using a different dataset (discussed Section 4.8) gives a range of streamflow depletion of between 321 and 332 L/s.

The results of the modelled scenarios provided a large range of results and are dependent on what input values are selected. The main differences in approaches that have resulted in the range of results are:

Aquifer conceptualisation - separating artesian wells and non-artesian wells versus excluding deep wells (>40 m) from the assessment

Calculation of seasonal allocation Aquifer parameters used in assessment

These assessments of streamflow depletion have highlighted the fact that there is a limited conceptual understanding of Lower Tukituki hydrogeology, and limited information on spatial distribution of aquifer parameters in the Lower Tukituki.

(10)

SKM agree with the approach of HBRC in working toward excluding ‘low risk’ groundwater takes, such as those in confined aquifer systems from the policy. However, SKM are not in agreement with the use of an arbitrary value of 40 m as used by HBRC in their modelling.

SKM recommend that the best approach to improve the accuracy of aquifer parameters and conceptualisation used in this modelling is to undertake further assessment of the geology, pump test data and groundwater flow lines to delineate individual aquifer boundaries. This would allow a more scientifically based decision on such approaches as the proposed cut-off depths to be made.

While a data intensive method, it is quite clear the current knowledge of aquifer dynamics in the Lower Tukituki is limited, and requires further refinement. This information would also allow a more accurate calculation of streamflow depletion to be undertaken.

SKM also recommend that if cut-off depth or similar approach is defined for the catchment, then this must be incorporated within the Policy in some form, perhaps as a technical report or guidance note. This information will help to inform applicants on the procedure that Council will undertake when determining the streamflow depletion effects of each individual groundwater take. Other guidance, such as clarification regarding the determination of seasonal irrigation volumes, could also be incorporated into this document.

Overall SKM are in general agreement with the proposed changes to policy outlined by both Ms Codlin and Mr Thorley, but as highlighted above, believe further revision is required once a more thorough assessment of streamflow depletion is undertaken. SKM also believe that clarification around the implementation of the policy, particularly in regards to the exclusion of users based on hydrogeological setting, would be crucial to applicants understanding of the policy implications on their groundwater takes.

(11)

1. Introduction

1.1. Purpose and Scope of Report

The purpose of this technical report is to present the results of work that has been undertaken to assess the potential scale of streamflow depletion effects in the Lower Tukituki Catchment as a result of currently consented groundwater abstractions. This work was completed as it has been identified by Horticulture New Zealand as a major data gap in the Plan Change, and Horticulture New Zealand have undertaken this work using a collaborative approach with Hawkes Bay Regional Council (HBRC).

The major data gap under the proposed changes to the plan stems from Policy TT11 under which all new and renewals of existing groundwater take consents will be required to assess their degree of connectivity (e.g. streamflow depletion as a percentage of abstraction rate) with the Tukituki River. Depending on the outcome of their assessment (compared against Table 5.9.7 – Low, Medium, High or Direct) the portion of the take that comes from the river can be counted as surface water allocation.

The current surface water allocation for the Tukituki River is proposed to be 1,072 L/s. It has been identified by Horticulture New Zealand and HBRC that the effect of potential streamflow depletion has not been fully taken into consideration when developing this surface water allocation limit. As such, this assessment investigates the total potential streamflow depletion rate from the currently consented abstractions, in order to provide an understanding as to whether or not the proposed surface water allocation limit is sufficient to incorporate these effects.

The work was commissioned by Horticulture New Zealand. The scenarios that have been

assessed are focused on the concerns of Horticulture New Zealand. The results of this assessment have been incorporated into Scenario modelling undertaken using SOURCE, as presented in the SOURCE Model Report.

The results have also been used to inform the economic evidence presented by Stuart Ford and Hamish Peacock’s planning evidence.

(12)

2. Tukituki Physical Setting

This section describes the natural environment in the vicinity of the Lower Tukituki River including surface water, geology and hydrogeology.

2.1. Surface Water Catchments

The Tukituki River Catchment is located in the southern part of the Hawke’s Bay region. The largest tributary of the Tukituki River is the Waipawa River. It joins the Tukituki River east of the Ruataniwha Plains and the Waipawa and Waipukurau townships, about half way down the catchment.

Other significant tributaries that cross the Ruataniwha Plains include the Mangaonuku Stream, Tukipo River and Makaretu Stream, Porangahau and the Maharakeke Streams.

Beneath the Ruataniwha Plains, a complex aquifer system of gravels, silts and clays contain water from rainfall and rivers. Water discharges the Ruataniwha Basin through springs, joining the Tukituki and Waipawa Rivers through their river beds.

Apart from some areas of exotic forestry, and the native vegetation in the Ruahine Forest Park, the catchment is largely deforested. Land use in the hill country is predominantly dry stock farming with more intensive farming on the plains, predominately horticulture.

The Lower Tukituki River is part of the iconic landscape of Te Mata Peak and the Kahuranaki range. The river is appreciated for its aesthetic, recreational and cultural values and with the aquifer, is valued for the water it provides for household, stock and public supply, commercial and irrigation users.

2.2. Geology

The Heretaunga Plains were formed during the last 250,000 years by river sediments deposited by the Tutaekuri, Ngaruroro and Tukituki rivers and coastal lagoon, estuarine and embayment

deposits. The fluvial deposits that accumulated as gravel river channels during and immediately after the glacial periods form permeable high yielding aquifers. The aquifers are interbedded with interglacial silt, clay, peat and shelly sand and clay down to explored depths of 250 m. There is a general layered structure with coarse permeable gravel beds alternating with fine impermeable beds (as shown on Figures 1, 2 and 3). The fine grained sediments separating gravel aquifers form aquicludes and aquitards, which can impede vertical groundwater flow (Dravid & Brown, 1997).

The permeable gravel beds form aquifers that reflect their formation as meandering river channels.

The Mohaka, Tukituki, Ngaruroro, Esk, and Tutaekuri rivers have undergone course adjustment in response to tectonic deformation. The tectonic influence has been a major influence on the

(13)

hydrology of the Hawke’s Bay Rivers and the deposition of the aquifer/aquiclude sequence of the Heretaunga depression (Dravid & Brown, 1997).

Figure 1. Topographical map of the Lower Tukituki catchment showing the locations of the geological cross-sections.

(14)

Figure 2. Pakipaki-Haumoana geological cross-section, redrawn from Dravid & Brown (1997).

(See A3 attachment at rear)

Figure 3. Karamu geological cross-section, redrawn from Dravid & Brown (1997).

(See A3 attachment at rear)

2.3. Hydrogeology

The Heretaunga basin has an area of 330 km² and has a maximum thickness of 400 m. It is made up of 7 primary aquifers including the Tukituki aquifer (Brooks, 2006). Where the Tukituki River intersects the Heretaunga Plains, it contributes water to a shallow semi-confined to confined aquifer system underlying its floodplain. The lower Tukituki groundwater resource is represented by a localised gravel aquifer deposited by the Tukituki River and is a semi confined aquifer with leaky conditions and high transmissivities in the range of 10,000 – 15,000 m²/day and a storativity of 0.015 (Dravid & Brown,1997). The average thickness of the Tukituki aquifer system is about 20 metres.

Groundwater data such as groundwater levels and aquifer tests suggest the lower Tukituki and Heretaunga Plains aquifer systems are hydraulically connected (Harper, 2013). The Tukituki aquifer is interbedded with, and overlies the main Heretaunga Plains aquifer system between Thompson Road and Karamu Stream (Figure 3). The western extent of the aquifer system is marked by the terrace east of Mangateretere – Havelock North Road. Along the coast, the beach gravel deposits from south Haumoana to Te Awanga merge with the Tukituki aquifer system (Figure 4) (Dravid & Brown,1997).

During summer increased groundwater abstraction for irrigation may cause piezometric pressures in the main aquifer system to decline below the level of the overlying Tukituki aquifer system. This reversal of the hydraulic gradient would produce a downward flow of groundwater from the Tukituki River aquifer system to recharge the underlying Heretaunga plains aquifer system (Dravid &

Brown, 1997).

(15)

Figure 4. Aerial extent of the aquifer systems on the Heretaunga Plains (from Dravid &

Brown (1997)).

Groundwater flows from the western Heretaunga Plains to the sea with a gradient of about 0.008 (Brooks, 2006). The piezometric contours measured across the Heretaunga Plains shows the Ngaruroro River as the major source of groundwater that flows eastwards towards Flaxmere and Hastings and the Tukituki River, then northeastwards toward Taradale and Napier, rather than directly to the coast. Piezometric surveys indicate a sudden drop in groundwater pressure from 10 to 9 m above sea level just north of Havelock North. This was hypothesised to have resulted from significant water loss from the aquifer through abstractions, leakage between aquifers or loss to surface water (Dravid & Brown, 1997).

(16)

Brooks (2006) developed a steady-state numerical groundwater model for the Heretaunga

groundwater basin. The main objectives for this model were to predict pumping effects on the basin water levels, predict effects of pumping wells on surface water, and determine the basin water balance. The simulation suggested that water levels have declined about 2 metres across the Heretaunga groundwater basin since groundwater pumping began in the early 1900s.

2.4. Groundwater Abstraction

There are 134 groundwater bores located in the Lower Tukituki, with a total of 113 groundwater take consents located in the Surface Water Allocation Zone 1 in the Proposed Plan. These consents are for a variety of water uses as outlined in Table 2-1.

Table 2-1: Water Use for Groundwater Take Consents in the Lower Tukituki

Consent Use Number of Consents

Crops 26

Orchards 53

Vineyard 11

Pastoral Farm 7

Private Facility 1

Residential 4

Public Water Supply 1

Nursery 1

Frost Protection 9

Many of the groundwater takes in the Lower Tukituki River Catchment are located on the Heretaunga Plains, where the Tukituki River leaves the valley to intersect the plains. The proposed plan changes reference groundwater take locations in relation to Black Bridge, Red Bridge and the Otane/Papanui Groundwater Basin for the purposes of low flow limit setting. The number of takes in each of this area is listed in Table 2-2 (excludes frost protection takes).

Table 2-2: Location of Lower Tukituki Groundwater Consents

Low Flow Limit Setting Location Number of Consents

Downstream of Black Bridge 9

Downstream of Red Bridge (but above Black Bridge)

79 Upstream of Red Bridge (but not in

the Otane Basin)

7

Otane Basin 9

(17)

3. Plan Change 6

The Hawke’s Bay Regional Resource Management Plan (including the Regional Policy Statement) is a second generation statutory planning document, and became operative in August 2006. It identifies regionally significant issues facing the region, and sets out regional level policies and region-wide rules for addressing those issues. Within the regional plan policy provisions, there are some catchment specific water allocation limits and minimum flows and some catchment specific water quality guidelines.

Plan Change 6 (Change 6) is the first of a number of catchment specific plan changes for the Hawke’s Bay region that seek to implement the National Policy Statement for Freshwater

Management, as well as address specific water allocation and water quality issues in the Tukituki catchment.

The principal features of Change 6 in relation to the connection between groundwater and surface water are outlined in Policy TT11 as outlined below. The Publically Notified version of Policy TT11 is presented first (in Section 3.1), and then the revised version submitted by Ms Helen Codlin of HBRC in her submission to the EPA is presented after (in Section 3.2).

3.1. Publically Notified Version of TT11

POL TT11 MANAGING GROUNDWATER TAKES HYDRAULICALLY CONNECTED TO SURFACE WATER BODIES

1. To generally manage the effects of groundwater takes on surface water bodies, including wetlands, in the following manner:

(a) An appropriate scientific method must be used by consent applicants to assess the depletion effect of the groundwater take on nearby surface water 36 e.g.: using Guidelines for the

Assessment of Groundwater Abstraction Effects on Stream Flow prepared by Environment Canterbury (Techniques for evaluating stream depletion effects, Supplement to the guidelines for the assessment of groundwater abstraction effects on stream flow (2000), Report No. R09/53, ISBN 978-1-86937-992-6);

(b) Subject to (a), the potential adverse effects of groundwater takes on surface water depletion shall be managed in accordance with Table 5.9.7;

(c) Groundwater takes that are classified as High or Medium in Table 5.9.7 shall be included within the surface water allocation limits described in POL TT8 and POL TT9;

(d) Groundwater takes that are classified as High in Table 5.9.7 shall be subject to the minimum flow limits in POL TT7 and POL TT9, provided that the predicted reduction in stream depletion that will arise from ceasing the groundwater take will occur within 10 days.

(18)

Table 5.9.7: Management of Surface Water Depletion Effects

Classification of surface water depletion effect

Magnitude of surface water depletion effect

Management approach

High The surface water depletion effect is

assessed as:

(a) 50% or greater of the average groundwater pumping rate37 after 100 days of pumping; and

(b) greater than 1 L/s.

The calculated loss of surface water is included in the surface water allocation regime, and specific minimum flow restrictions are imposed on the groundwater take, subject to the proviso in POL TT11(1)(d).

Medium The surface water depletion effect is

assessed as:

(a) 20% or greater and less than 50%

of the average groundwater pumping rate34 after 100 days of pumping; and (b) greater than 1 L/s.

The calculated loss of surface water is included in the surface water allocation regime, but no specific minimum flow restrictions are imposed on the groundwater take.

Low The surface water depletion effect is

assessed as:

(a) less than 20% of the average groundwater pumping rate after 100 days of pumping; or

(b) 1 L/s or less.

The calculated loss of surface water is not included in the surface water allocation regime, and no specific minimum flow restrictions are imposed on the groundwater take.

3.2. Policy TT11 Version Submitted by Ms Codlin (HRBC)

POL TT11 MANAGING GROUNDWATER TAKES HYDRAULICALLY CONNECTED TO SURFACE WATER BODIES

1. To generally assess the effects of groundwater takes on surface water bodies, including wetlands, in the following manner:

(a) An initial assessment can be based on a review of well locations, water levels and well lithology records, and the use of an appropriate scientific model using existing or known transmissivity and storativity values to determine whether surface water depletion is likely to be a concern and estimate the potential surface water depletion effects.

(c) In the event that reliable data are not available to make the initial assessment, the applicant will be required to undertake an independent assessment of stream depletion effects using an

appropriate scientific method e.g. using Guidelines for the Assessment of Groundwater Abstraction Effects on Stream Flow prepared by Environment Canterbury R00/11, ISBN 1-86937-387-1 and Techniques for evaluating stream depletion effects, Supplement to the guidelines for the

assessment of groundwater abstraction effects on stream flow (2000), Report No. R09/53, ISBN 978-1-86937-992-6);

2. To generally manage the effects of groundwater takes on surface water bodies, including wetlands, in the following manner:

a) The potential adverse effects of groundwater takes on surface water depletion shall be managed in accordance with Table 5.9.7;

(19)

b) Groundwater takes that are classified as Direct, High or Medium in Table 5.9.7 shall be included within the surface water allocation limits described in POL TT8 and POL TT9;

c) Groundwater takes that are classified as Direct in Table 5.9.7 shall be subject to the minimum flow limits in POL TT7 and POL TT9, provided that the predicted reduction in stream depletion that will arise from ceasing groundwater take will occur within 10 days;

d) Groundwater takes that are classified as High in Table 5.9.7 shall be subject to rate of take / volume restrictions as described in POL TT9, except that irrigation takes shall be able to continue to take up to 50% of the daily volume as specified in their consent conditions for the period when flows are at or below the minimum flow limit and provided that the predicted reduction in stream depletion that would arise from ceasing groundwater take would occur within 10 days.

Table 5.9.7: Management of Surface Water Depletion Effects

Classification of surface water depletion effect

Magnitude of surface water

depletion effect Management approach

Direct The surface water depletion is

assessed as:

(a) 90% or greater of the average groundwater pumping rate2 after 150 days of pumping; and (b) greater than 1 L/sec

The calculated loss of surface water is included in the surface water allocation regime and specific minimum flow restrictions are imposed on the groundwater take, subject to the proviso in POL TT11(2)(c).

High The surface water depletion effect is

assessed as:

(c) 60% or greater and less than 90% of the average groundwater pumping rate3 after 150 days of pumping; and

(d) greater than 1 L/s.

The calculated loss of surface water is included in the surface water allocation regime and specific rate of take / volume restrictions are imposed on the take in accordance with POL TT9 and POL TT11(2)(d).

Medium The surface water depletion is

assessed as:

(a) 20% or greater and less than 60% of the average groundwater pumping rate3 after 150 days of pumping; and

(b) greater than 1 L/s.

The calculated loss of surface water is included in the surface water allocation regime, but no specific minimum flow, rate of take or volume restrictions are imposed on the groundwater take.

Low The surface water depletion effect is

assessed as:

(a) less than 20% of the average groundwater pumping rate after 150 days of pumping; or (b) 1 L/s or less.

The calculated loss of surface water is not included in the surface water allocation regime and no specific minimum flow or rate of take restrictions are imposed on the groundwater take.

2 The average groundwater pumping rate is based on the seasonal or annual volume averaged over 150 days or full year whichever is applicable assuming pumping occurs for 24 hours per day

3 The average groundwater pumping rate is based on the seasonal or annual volume averaged over 150 days or full year whichever is applicable assuming pumping occurs for 24 hours per day

3.3. Proposed Amendments to Policy TT11

Mr Michael Thorley has presented evidence on behalf of HBRC and provides an assessment of Policy TT11. In Section 9 of his evidence he outlines his further recommendations for this Policy.

(20)

Recommending that HBRC clarifies provision in Policy TT11.2 (c) and (d) by inserting the word “any”, so that it now reads “provided that any predicted reduction in stream depletion that will arise from ceasing groundwater take will occur within 10 days

Recommending that the 1 L/s threshold be amended to 2 L/s. This recommendation has been presented as Mr Thorley states “In my opinion, 1 L/s is a low threshold, and may be overly conservative when compared to the effect such a take might have on surface water bodies. I suggest that it be amended to 2 L/s, consistent with policy and practice in Canterbury and Southland” (paragraph 2.12).

3.4. Version adopted for this assessment

The stream depletion assessments presented in this report are analysed using the version of TT11 presented by Ms Codlin in her evidence. In addition, the assessments test the recommendation from Mr Thorley regarding increasing the 1 L/s threshold to 2 L/s.

(21)

4. Methodology

4.1. Overview

Groundwater and surface water systems can be connected, and groundwater discharge can often be a significant component of the flow in a surface water body (i.e. baseflow). Pumping of

groundwater can lead to a reduction in the amount of groundwater that flows into a stream. Stream depletion may occur if a stream/river is:

receiving groundwater flow (i.e. nearby groundwater levels are higher than stream water levels);

if it is in equilibrium with groundwater (i.e. stream levels and groundwater levels are equal); or in some cases if a stream is losing flow to groundwater (i.e. groundwater levels are lower than stream levels) (Environment Canterbury Report R00/11).

There are several factors that determine the degree of connection between groundwater and surface water including:

the hydraulic conductivity of the strata between the stream and pumping bore screen location;

the depth of the production bore, as with increasing production bore depth, the likelihood for stream depletion decreases due to the intervening low permeability units above the screened water bearing unit.

HRBC have proposed to manage streamflow reduction by classifying the streamflow depletion portion of a groundwater take as surface water, and requiring this to come from the surface water allocation of the Tukituki River. Depending on the degree of connectivity, flow restrictions may be placed on groundwater takes when the Tukituki River reaches its minimum flow level at a particular reference point as outlined in Table 5.9.7 of POL TT11)

There are numerous quantitative tools and analytical methods that can be used to calculate the potential streamflow depletion as a result of groundwater abstraction that range from high level screening tools to complex analytical methods involving considerable data input. As stated in Environment Canterbury (2000), these tools and methods are a gross simplification of the complex variability that exists in naturally deposited groundwater systems, however these methods are considered to be the most appropriate means of estimating stream depletion effects.

4.2. SKM Stream Depletion Calculations

In order to assess the cumulative impacts of streamflow depletion, SKM completed a high level analytical modelling exercise using Glover’s Analytical Equation and Jenkins Stream Depletion Factor (SDF). These methods require minimal data input and provide a screening tool that can be

(22)

The Glover’s Analytical Solution (Glover and Balmer, 1954) is based on a highly simplified stream- aquifer-well system. The Glover solution provides a total rate of streamflow depletion as a function of time and is a product of the pumping rate of the well, Qw, and a mathematical function referred to as the complementary error function, erf(z):

QS = QW erfc(z)

Variable z is equal to /(4 ), where d is the distance between the well and the stream, S is the storage coefficient of the aquifer (specific yield for water table aquifers), T is the transmissivity of the aquifer, and t is the time (USGS, 2012).

Jenkins (1968), defined the quantity d²/D, which is the equivalent to (d²S)/(T) in the above equation as the Stream Depletion Factor (SDF). The SDF has units of time, such as days, and is essentially the number of days before the pumping well starts to draw from the river.

4.3. HBRC Stream Depletion Calculations

Additional modelling undertaken by HBRC Consultant Groundwater Specialist Mr Thorley uses the Hunt (2003) solution for stream depletion.

The Hunt (2003) solution described in Hunt (2008) aims to predict stream depletion due to

abstraction where the stream partially penetrates an aquitard that overlies a permeable aquifer unit from which abstraction takes place. This setup was developed from a pump test model known as the Boulton model, where drawdown is predicted in an aquifer unit overlain by an aquitard

containing the water table. The Hunt (2003) model assumes that the aquifer is infinite and that flow is only able to move vertically in the aquitard layer and horizontally in the aquifer layer as in the Boulton model (as shown in Figure 5).

To use Hunt (2003), the pumping bore must be at least 10 stream widths away from the stream and the stream must be located in the aquitard unit, which should have a significantly lower

horizontal conductivity than the pumped aquifer unit. The aquitard unit can be used to model layers of different geology under the condition that the top layer that is penetrated by the stream is an aquitard layer, preventing significant horizontal flow to or from the stream.

(23)

Figure 5: Hunt (2003) stream depletion schematic.

4.4. Data Sources

The following groundwater consent and well information was supplied by HBRC and was used in the assessment of streamflow depletion:

A Well Database search covering all of the wells that fall within the Surface Water Allocation Zone 1, and the Otane Basin/Papanui catchment. This was provided by Simon Harper of HBRC on the 27^th September, 2013. The wells that are located just outside the boundary of Surface Water Allocation Zone 1 in the Lower Tukituki are excluded from this assessment.

A copy of all groundwater and surface water takes within Surface Water Allocation Zone 1.

This was provided by Paul Barrett of HBRC on the 27th September 2013. Groundwater takes recorded as ‘Frost Protection’ were not included in the assessment as it is assumed that they will not be abstracting during the normal irrigation season or in times of low flow.

Seasonal allocations for all groundwater consents calculated by HBRC using crop water requirements and the irrigation area for the consent. These allocations were provided by Paul Barrett of HBRC on the 2nd October 2013.

4.5. Input parameters

4.5.1. Transmissivity and Aquifer Storage

The streamflow depletion calculations outlined above require inputs of transmissivity and storage coefficient or specific yield. HBRC only hold limited aquifer test information on wells in the Lower Tukituki, however a review of available data provided the range of values presented in Table 4-1. It should be noted that all of these aquifer tests were undertaken on semi-confined or confined wells.

(24)

Table 4-1: Summary of Aquifer Test Parameters

Bore ID Depth (m) Distance (m)

Transmissivity

(m²/day) S

5830 36.5 - 950 1.50E^-04

5759 95 942.6 9000 1.00E^-05

435 39 - 1076 1.80E^-05

435 39 - 1174 4.30E^-04

5009 44 - 757 3.00E^-04

5009 44 - 3160

9049 16 1146.5 12200 1.60E^-03

1137 12.5 - 1300 2.00E^-02

15158 36.5 1108.5 1000 4.00E^-04

The model parameters that were determined to be used in groundwater depletion scenarios are outlined in Table 4-2. These parameters have been discussed and agreed with HBRC’s

groundwater expert, Mr Thorley, as being appropriate for this high level assessment.

Table 4-2: Aquifer Test Parameters Used in Assessments

Model Parameters Unconfined well values

Confined well values

T (transmissivity) 1200 m²/day 1200 m²/day

Sy (specific yield) 0.1 -

S (storage coefficient) - 0.0001

t (time) 150 days 150 days

4.5.2. Distance to the River

The stream depletion calculations require the distance from the groundwater take to the stream.

This distance has been calculated using the coordinates from the consent list and the nearest surface water body – this could be the main branch of the Tukituki, or a tributary – the closest of the two was used in the assessment.

4.5.3. Seasonal Allocation

The stream depletion calculations require the input of Qw, the pumping rate of the well. Policy TT11 specifies seasonal allocations for the assessment, and as the annual allocation volumes for each groundwater take calculated by HBRC were used in this assessment. In the first instance,

(25)

HBRC has used the consented annual volume, however in the event the consent does not have an annual volume associated with it, the annual irrigation volumes based on crop water requirements has been divided by 150 to give a daily abstraction. This value was then compared with either the daily or weekly consented volume, and the lesser value was used in the analysis.

4.5.4. Aquifer Confinement

Conceptually, the deeper the well, the less likely it is to have an influence on streamflow depletion as confining layers will reduce the connectivity between the screened section of the well and a surface water body. Determining the degree of confinement of each borehole is a time consuming and data hungry process that requires a detailed review of borelogs and water levels. For this initial high level assessment of streamflow depletion, SKM has classified wells as confined and unconfined based on the ‘Artesian’ category in the HBRC Wells database. This category lists wells that are confined as ‘Flowing Artesian’ or ‘Non-Flowing Artesian’.

4.5.5. Multiple Wells per Consent

There are several consents that abstract the full consented volume from multiple wells. In the case where consent is to abstract from multiple wells, the abstraction volumes have been averaged across each well. This method was collaboratively agreed with HBRC as the best approach for this assessment.

4.6. SKM Modelled Scenarios

Using the above data, two scenarios have been modelled by SKM:

1) Assuming wells listed in the HBRC database as ‘Flowing or Non-Flowing Artesian’ as confined and all remaining as unconfined, and that the cut-off for L is <1 L/s

2) Assuming wells listed in the HBRC database as ‘Flowing or Non-Flowing Artesian’ as confined and all remaining as unconfined, and that the cut-off for L is <2 L/s, as per the evidence of the HBRC Groundwater Specialist Mr Thorley.

The results of this modelling are presented in Section 5, and are compared to the modelling undertaken by HBRC.

4.7. Comparison to Plan Change Table 5.9.7 Criteria

Using the stream depletion rate calculated by Glovers and Jenkins, the ratio of stream depletion (L/s) to consented abstraction rate can be compared. This ratio allows the stream depletion rate to be classified against Table 5.9.7 and gives each take a Low (L), Medium (M), High (H) or Direct (D) Category.

(26)

By summing the totals that fall within each M, H or D category, a total stream depletion rate (L/s) for the Lower Tukituki can be calculated for each of the Scenarios listed in Section 4.6.

It is these modelled depletion rates that can be used as an initial assessment of whether there is sufficient allocation in the proposed Plan Change to allow for these groundwater users.

4.8. HBRC Streamflow Depletion Modelling

Using the Jenkins streamflow depletion assessment undertaken by SKM, HBRC has undertaken further assessments using the Hunt (2003) methodology (described in Section 4.3).

The Hunt (2003) solution has the capability of assessing stream depletion in more detail through the use of a streambed conductance value ( ) and aquifer thickness. However in the absence of site specific data, returns a very similar result to Jenkins.

HBRC have used a number of assumptions that are not consistent with the SKM modelling, and therefore a range of streamflow depletion results are presented in Section 5. The main differences between the two approaches are:

1) HBRC have used an arbitrary value of 40 m as their cut-off for a confined well. That is, it is assumed that any well below 40m is confined, and has been excluded from the streamflow depletion calculations

2) HBRC have taken the lowest of consented daily abstraction rate, weekly maximum volume, 28 day maximum volume, consented annual volume or calculated crop water requirements as their abstraction rate (Qw) for the calculations.

The HBRC streamflow depletion assessment uses the values in Table 4-2.

Table 4-2: HBRC Hunt (2003) Stream Depletion Input Values

Q (m³/day)

T

(m²/day) S K’/B’ (1/day) Sy (m/day)

1200 0.1 10000000000 0.1 1E⁺¹²

(27)

5. Results

The following section summarises the results of the six modelled scenarios and totals the stream flow depletion for each scenario.

5.1. Scenario 1: SKM - Calculating artesian and unconfined separately and L = <1 L/s

Scenario 1 used the confined well values from Table 4-2 for ‘Flowing or Non-Flowing Artesian’ well and unconfined values for all other wells. Each groundwater take was then classified as direct, high, medium or low using Table 5.9.7 but assuming that the limit on stream depletion is up to 1 L/s (Table 5-1 and 5-2). These were then totalled in Table 5-3. Total stream depletion from

groundwater takes is 53,566 m³/day for Scenario 1.

Table 5-1. Confined Bores Only

Table 5.9.7.

Classification No. SFD (L/s) SFD (m³/day)

L 8

M 1 2 177

H 46 144 12,437

D 22 285 24,661

Total Depletion 431 37,275

Table 5-2. Unconfined Bores Only

Table 5.9.7.

L 16

M 15 53 4,558

H 2 14 1,228

D 8 122 10,505

Total Depletion 189 16,291

Table 5-3. Total of Confined and Unconfined Bores

Table 5.9.7.

L 24

M 16 55 4,735

H 48 158 13,665

D 30 407 35,166

Total Depletion 620 53,566

(28)

5.2. Scenario 2: SKM - Calculating artesian and unconfined separately and L = <2 L/s

Scenario 2 used the confined well values from Table 4-2 for ‘Flowing or Non-Flowing Artesian’ well and unconfined values for all other wells. Each groundwater take was then classified as direct, high, medium or low using Table 5.9.7 but assuming that the limit on stream depletion is up to 2 L/s as per the evidence of Mr Thorley (Table 5-4 and Table 5-5). These were then totalled in Table 5-6. Total stream depletion from groundwater takes is 50,733 m³/day for Scenario 2.

Table 5-4. Confined Bores Only

Table 5.9.7.

L 30

M 1 2 177

H 27 116 10,014

D 19 281 24,251

Total Depletion 399 34,442

Table 5-5. Unconfined Bores Only

Table 5.9.7.

L 16

M 15 53 4,558

H 2 14 1,228

D 8 122 10,505

Total Depletion 189 16,291

Table 5-6. Total of Confined and Unconfined Bores

Table 5.9.7.

L 46

M 16 55 4,735

H 29 130 11,242

D 27 403 34,756

Total Depletion 588 50,733

(29)

5.3. Scenario 3: HBRC - Jenkins, Low = <1 L/s

Scenario 3 was modelled by HBRC and models the stream depletion rates using the Jenkins solutions where the ‘Low’ cutoff as per Table 5.9.7 is < 1 L/s. The dataset of groundwater takes is different from that used in the SKM modelling results, as discussed in Section 4.8. The results for the HBRC Jenkins assessment are presented in Table 5-7. Table 5-7. HBRC Stream Depletion using Jenkins

Table 5-7. HBRC Stream Depletion using Jenkins, L=<1 L/s

Table 5.9.7.

L 28

M 21 71 6,145

H 35 144 12,443

D 5 146 12,634

Total Depletion 361 31,223

5.4. Scenario 4: HBRC - Hunt (2003), Low = <1 L/s

Scenario 4 models the stream depletion rates using the Hunt (2003) solution where the ‘Low’ cutoff as per Table 5.9.7 is < 1 L/s. The dataset of groundwater takes is different from that used in the SKM modelling results, as discussed in Section 4.8. The results for the HBRC Hunt (2003) are presented in Table 5-8.

Table 5-8. HBRC Stream Depletion using Hunt (2003), L=<1 L/s

Table 5.9.7.

L 21

M 9 23 1,986

H 45 133 11,510

D 14 211 18,189

Total Depletion 367 31,686

5.5. Scenario 5: HBRC - Jenkins, Low = <2 L/s

HBRC Scenario 5 compares stream depletion rates using the Jenkins solutions where the ‘Low’

cutoff as per Table 5.9.7 is < 2 L/s. The dataset of groundwater takes is different from that used in the SKM modelling results, as discussed in Section 4.8. The results for the HBRC Jenkins assessment are presented in Table 5-9.

(30)

Table 5-9. HBRC Stream Depletion using Jenkins, L=<1 L/s

Table 5.9.7.

L 47

M 16 63 5,435

H 21 123 10,616

D 5 146 12,634

Total Depletion 332 28,685

5.6. Scenario 6: HBRC – Hunt (2003), Low = <2 L/s

Scenario 6 models the stream depletion rates using the Hunt (2003) solutions where the ‘Low’

cutoff as per Table 5.9.7 is < 2 L/s. The dataset of groundwater takes is different from that used in the SKM modelling results, as discussed in Section 4.8. The results for the HBRC Hunt (2003) are presented in Table 5-10.

Table 5-10. HBRC Stream Depletion using Hunt (2003), L=<2 L/s

Table 5.9.7.

L 50

M 7 20 1,722

H 21 96 8,272

D 11 205 17,753

Total Depletion 321 27,747

(31)

6. Discussion

Currently the surface water allocation in the Lower Tukituki (Schedule XVIII Table) allows for all surface water takes, and all groundwater takes categorised as Stream Depleting. The groundwater takes currently categorised as ‘Stream Dep’ have been categorised as such based on historical pump test data and a variety of methodologies. Using this historical categorisation, it has been determined by HBRC that the portion of surface water being abstracted from the Tukituki River from groundwater takes is approximately 312 L/s.

It has been acknowledged that the methods used to determine the Stream Depleting groundwater bores may not be appropriate, and as such the 312 L/s included in the surface water allocation may not accurately reflect the level of effect from the current groundwater users in this catchment. As such, assessments were completed by SKM and HBRC to investigate the total potential streamflow depletion rate from the currently consented abstractions, and as such provide an understanding as to whether or not the proposed surface water allocation limit is sufficient to incorporate these effects.

As outlined in Section 5, SKM and HBRC have modelled six different streamflow depletion scenarios. The different methodologies, and different input parameters used by SKM and HBRC have results in a range of stream depletion rates of between 321 and 588 L/s.

It is clear from the results of the scenarios presented above that these high level methods of determining streamflow depletion provide a large range of results and are dependent on what input values are selected. The main differences in approaches that have resulted in the range of results are:

1) Aquifer conceptualisation - separating artesian wells and non-artesian wells versus excluding deep wells (>40 m) from the assessment

2) Calculation of seasonal allocation 3) Aquifer parameters used in assessment These three issues are discussed in detail below.

6.1. Aquifer Conceptualisation

There has been two separate approaches used by SKM and HBRC when determining the aquifer conceptualisation for the assessments. SKM have used a high level approach of assessing the wells based on whether or not they have been identified as having artesian flow. In comparison, HBRC has determined a cutoff depth for determining confinement of a well. It is SKM’s

understanding that this arbitrary cutoff was determined following a high level review of the geology in the vicinity of the Lower Tukituki.

(32)

As outlined in Section 2.2, the aquifers in the vicinity of the Lower Tukituki River are interbedded with interglacial silt, clay, peat and shelly sand and clay down to explored depths of 250 m. There is a general layered structure with coarse permeable gravel beds alternating with fine impermeable beds. The fine grained sediments separating gravel aquifers form aquicludes and aquitards, which can impede vertical groundwater flow.

While SKM acknowledge that the high level assessment completed using artesian and non- artesian wells is highly conservative based on this aquifer conceptualisation, we also do not support the approach suggested by HBRC, as there is no clear geological justification for the determination of 40 m.

In paragraph 8.3 of his evidence, Mr Thorley states “I consider that regional investigations to establish some of the key inputs to hydraulic connection assessments, such as stream bed

conductance, would be useful. Whether such investigations are carried out by HBRC is a matter for HBRC to consider.” SKM believes that these assessments are required to be completed in order to accurately determine the level of streamflow depletion and to provide certainty to the current consent holders in the catchment that the current level of effects have been incorporated or considered within the plan. It can be clearly seen, based on the large range of stream depletion rates calculated in the six scenarios in Section 5, what effect different methods of incorporating geology into the streamflow depletion calculations can have on the overall depletion effect.

Overall, SKM believe that the best approach would be to undertake further assessment of the geology, pump test data and groundwater flow lines to delineate individual aquifer boundaries.

This would allow a more scientifically based decision on such approaches as the proposed cut-off depths to be made. While a data intensive method, it is quite clear the current knowledge of aquifer dynamics in the Lower Tukituki is limited, and requires further refinement.

In addition to these assessments, SKM also believe that if a cutoff depth or similar approach is defined for the catchment, then this must be incorporated within the Policy in some form, perhaps as a technical report or guidance note. This information will help to inform applicants on the procedure that Council will undertake when determining the streamflow depletion effects of each individual groundwater take.

6.2. Calculation of Seasonal Allocation

The footnotes of Policy TT11 provide guidance on how to calculate the average groundwater pumping rate required to determine the classification in Table 5.9.7, e.g. “the average groundwater pumping rate is based on the seasonal or annual volume averaged over 150 days or full year whichever is applicable assuming pumping occurs for 24 hours per day”.

Although these footnotes appear to provide guidance to the applicants regarding how to assess their groundwater takes, there does appear to be additional variations to how these footnotes are interpreted, and ultimately these do have an effect on the streamflow depletion effects that are calculated.

(33)

These variations stem from the fact that the majority of the currently consented groundwater takes do not have a consented seasonal allocation. As such, for the purposes of this assessment, these were calculated by HBRC. SKM incorporated these seasonal allocations into their streamflow depletion assessment, while HBRC completed an additional assessment through comparing the seasonal allocation (converted into a daily rate) with daily rates calculated from consented weekly rates. They then utilised the lesser daily rate for the streamflow assessment.

SKM believe that, as with the cutoff depths, this information must be incorporated within the Policy in some form, perhaps as a technical report or guidance note. This information will help to inform applicants on the procedure that Council will undertake when determining the streamflow depletion effects of each individual groundwater take.

6.3. Aquifer Parameters

As stated in Section 4.5.1, HBRC only hold limited aquifer test information on wells in the Lower Tukituki. In addition, a lot of this data has not been independently reviewed by HBRC so its accuracy is unknown. The assessments outlined in Section 5 have all been undertaken using single aquifer parameters, although a range could have been incorporated. The methods used to determine stream depletion are sensitive to the aquifer parameters used and as such undertaking additional assessments based on varying aquifer parameters would only have further highlighted additional uncertainty in the calculated stream depletion values presented.

In addition, it should be noted that there is very little difference between the resulting streamflow depletions calculated by HBRC using Jenkins, and those made using Hunt (2003). This is to be expected as the Hunt solution is based on Jenkins, but has been refined to include a value (streambed conductance). In the absence of any streambed conductance information from the Lower Tukituki, the HBRC modelled scenarios use a high , which in effect assumes that the stream is well connected to the groundwater (no resistance to flow).

Once again, this process has highlighted that there is only limited data available on which to base these calculations and that further work is required by HBRC to enable more accurate assessments of streamflow depletion to be made.

(34)